Symfony Profiler

{
    "@context": [
        "https:\/\/join-lemmy.org\/context.json",
        "https:\/\/www.w3.org\/ns\/activitystreams"
    ],
    "actor": "https:\/\/lemmy.world\/c\/selfhosted",
    "to": [
        "https:\/\/www.w3.org\/ns\/activitystreams#Public"
    ],
    "object": {
        "id": "https:\/\/discuss.tchncs.de\/activities\/create\/78afbfa6-45d5-490a-82d2-eff314878b8e",
        "actor": "https:\/\/discuss.tchncs.de\/u\/Natanox",
        "@context": [
            "https:\/\/join-lemmy.org\/context.json",
            "https:\/\/www.w3.org\/ns\/activitystreams"
        ],
        "to": [
            "https:\/\/www.w3.org\/ns\/activitystreams#Public"
        ],
        "object": {
            "type": "Note",
            "id": "https:\/\/discuss.tchncs.de\/comment\/17486775",
            "attributedTo": "https:\/\/discuss.tchncs.de\/u\/Natanox",
            "to": [
                "https:\/\/www.w3.org\/ns\/activitystreams#Public"
            ],
            "cc": [
                "https:\/\/lemmy.world\/c\/selfhosted",
                "https:\/\/lemmy.dbzer0.com\/u\/marauding_gibberish142"
            ],
            "content": "<p>I\u2019m currently looking for this as well. As far as my investigation went right now I\u2019ll probably go for 2x AMD Instinct MI50. Each of them has equivalent to slightly higher performance than a P40, however usually only 16gb VRAM (If you\u2019re super lucky you might get one with 32gb, those are usually not labeled as such though; probably binned MI60). With two of them you got 32gb VRAM and quite the performance for, right now, 200\u20ac \/ card. Alternatively you should be able to run quantized models on a single card as well.<\/p>\n<p>If you don\u2019t mind running ROCm instead of CUDA this seems like a good bang for the buck. Alternatively you might look into AMDs new line of \u201cAI\u201d SoCs (for example Frameworks Desktop computer). They seem to be really good as well, and depending on your usecase might be more useful than an equally priced 4090.<\/p>\n",
            "inReplyTo": "https:\/\/lemmy.dbzer0.com\/post\/41292446",
            "mediaType": "text\/html",
            "source": {
                "content": "I'm currently looking for this as well. As far as my investigation went right now I'll probably go for 2x AMD Instinct MI50. Each of them has equivalent to slightly higher performance than a P40, however usually only 16gb VRAM (If you're super lucky you might get one with 32gb, those are usually not labeled as such though; probably binned MI60). With two of them you got 32gb VRAM and quite the performance for, right now, 200\u20ac \/ card. Alternatively you should be able to run quantized models on a single card as well.\n\nIf you don't mind running ROCm instead of CUDA this seems like a good bang for the buck. Alternatively you might look into AMDs new line of \"AI\" SoCs (for example Frameworks Desktop computer). They seem to be really good as well, and depending on your usecase might be more useful than an equally priced 4090.",
                "mediaType": "text\/markdown"
            },
            "published": "2025-04-02T11:42:09.359030Z",
            "tag": [
                {
                    "href": "https:\/\/lemmy.dbzer0.com\/u\/marauding_gibberish142",
                    "name": "@marauding_gibberish142@lemmy.dbzer0.com",
                    "type": "Mention"
                }
            ],
            "distinguished": false,
            "language": {
                "identifier": "en",
                "name": "English"
            },
            "audience": "https:\/\/lemmy.world\/c\/selfhosted",
            "attachment": []
        },
        "cc": [
            "https:\/\/lemmy.world\/c\/selfhosted",
            "https:\/\/lemmy.dbzer0.com\/u\/marauding_gibberish142"
        ],
        "tag": [
            {
                "href": "https:\/\/lemmy.dbzer0.com\/u\/marauding_gibberish142",
                "name": "@marauding_gibberish142@lemmy.dbzer0.com",
                "type": "Mention"
            }
        ],
        "type": "Create",
        "audience": "https:\/\/lemmy.world\/c\/selfhosted"
    },
    "cc": [
        "https:\/\/lemmy.world\/c\/selfhosted\/followers"
    ],
    "type": "Announce",
    "id": "https:\/\/lemmy.world\/activities\/announce\/create\/2d3f11d5-4dd0-450c-b9fd-49cf4d4cb233"
}

{"@context":["https://join-lemmy.org/context.json","https://www.w3.org/ns/activitystreams"],"actor":"https://lemmy.world/c/selfhosted","to":["https://www.w3.org/ns/activitystreams#Public"],"object":{"id":"https://discuss.tchncs.de/activities/create/78afbfa6-45d5-490a-82d2-eff314878b8e","actor":"https://discuss.tchncs.de/u/Natanox","@context":["https://join-lemmy.org/context.json","https://www.w3.org/ns/activitystreams"],"to":["https://www.w3.org/ns/activitystreams#Public"],"object":{"type":"Note","id":"https://discuss.tchncs.de/comment/17486775","attributedTo":"https://discuss.tchncs.de/u/Natanox","to":["https://www.w3.org/ns/activitystreams#Public"],"cc":["https://lemmy.world/c/selfhosted","https://lemmy.dbzer0.com/u/marauding_gibberish142"],"content":"<p>I’m currently looking for this as well. As far as my investigation went right now I’ll probably go for 2x AMD Instinct MI50. Each of them has equivalent to slightly higher performance than a P40, however usually only 16gb VRAM (If you’re super lucky you might get one with 32gb, those are usually not labeled as such though; probably binned MI60). With two of them you got 32gb VRAM and quite the performance for, right now, 200€ / card. Alternatively you should be able to run quantized models on a single card as well.</p>\n<p>If you don’t mind running ROCm instead of CUDA this seems like a good bang for the buck. Alternatively you might look into AMDs new line of “AI” SoCs (for example Frameworks Desktop computer). They seem to be really good as well, and depending on your usecase might be more useful than an equally priced 4090.</p>\n","inReplyTo":"https://lemmy.dbzer0.com/post/41292446","mediaType":"text/html","source":{"content":"I'm currently looking for this as well. As far as my investigation went right now I'll probably go for 2x AMD Instinct MI50. Each of them has equivalent to slightly higher performance than a P40, however usually only 16gb VRAM (If you're super lucky you might get one with 32gb, those are usually not labeled as such though; probably binned MI60). With two of them you got 32gb VRAM and quite the performance for, right now, 200€ / card. Alternatively you should be able to run quantized models on a single card as well.\n\nIf you don't mind running ROCm instead of CUDA this seems like a good bang for the buck. Alternatively you might look into AMDs new line of \"AI\" SoCs (for example Frameworks Desktop computer). They seem to be really good as well, and depending on your usecase might be more useful than an equally priced 4090.","mediaType":"text/markdown"},"published":"2025-04-02T11:42:09.359030Z","tag":[{"href":"https://lemmy.dbzer0.com/u/marauding_gibberish142","name":"@marauding_gibberish142@lemmy.dbzer0.com","type":"Mention"}],"distinguished":false,"language":{"identifier":"en","name":"English"},"audience":"https://lemmy.world/c/selfhosted","attachment":[]},"cc":["https://lemmy.world/c/selfhosted","https://lemmy.dbzer0.com/u/marauding_gibberish142"],"tag":[{"href":"https://lemmy.dbzer0.com/u/marauding_gibberish142","name":"@marauding_gibberish142@lemmy.dbzer0.com","type":"Mention"}],"type":"Create","audience":"https://lemmy.world/c/selfhosted"},"cc":["https://lemmy.world/c/selfhosted/followers"],"type":"Announce","id":"https://lemmy.world/activities/announce/create/2d3f11d5-4dd0-450c-b9fd-49cf4d4cb233"}